Creativity in Schizophrenia: Evidence Beyond Anecdotes

ABSTRACT

Recent meta-analytic work has highlighted lower creativity in schizophrenic patients as compared to control participants, but the cognitive and affective underpinnings of this difference still need to be fully understood. To this purpose, we adopted a multi-measure approach and compared a group of schizophrenic outpatients and a group of demographically-matched controls (N = 34) on the Alternative Uses Test (AUT) and the Remote Association Test (RAT), assessing divergent and convergent thinking respectively. The participants’ cognitive status, affective status, and response inhibition skills were also appraised. The results showed a significantly worse performance in patients vs. control participants in both tests of creativity. Performance was also negatively correlated with patients’ symptoms, being lower in more severe clinical conditions. The difference between groups in the RAT was no more significant when controlling for individuals’ cognitive functioning, in line with previous studies in healthy populations. In contrast, the difference in AUT fluency remained significant even when controlling for cognitive and affective measures. Our findings suggest that creativity reduction in schizophrenic patients involves both convergent and divergent thinking, and that the latter aspect is not simply the consequence of a general cognitive or inhibitory impairment.

Acknowledgments

The authors thank Vittorio Bortolin and Irene Teresa Florean for their help during data collection and scoring. The findings in the manuscript have not been published elsewhere, even if there is overlap between the predictor variables and the sample used in the present study and the ones employed in a previous study we reported (Del Missier, F., Galfano, G., Venerus, E., Ferrara, D., Bruine De Bruine, W., & Penolazzi, B. (2020). Decision-making competence in schizophrenia. Schizophrenia Research, 215, 457-459). In particular, although the sample and the predictor variables used in the present study are the same as the ones employed in the previous study, the aims of the present study, the criterion variables, the analyses, and the results are entirely different, given that the present study focuses on creativity whereas the previous one focused on decision-making competence.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability

Data are available from the corresponding author upon reasonable request.

Notes

1. Several definitions of inhibition have been proposed and inhibitory processes have been investigated with a variety of measures (see e.g., Friedman & Miyake, 2004). In this study, in line with the majority of studies on inhibition in schizophrenia, we focus on response inhibition, which can be defined as the ability to withhold a prepotent response, and is usually assessed by the Stroop task, the Sustained Attention Response Task (SART), or the antisaccade task.

2. In the present study, we did not employ an originality score based on uniqueness of responses, given that this measure is sensitive to sample size (and thus potentially problematic in clinical studies). Moreover, in the very few cases in which a participant did not produce a single valid response for a stimulus object (7 cases overall), his/her score for that object on the originality and feasibility variables was 1 (i.e., the lowest point of the scales). Indeed, not producing any alternative use means failing to produce even a minimally original and feasible response. We also employed a more lenient alternative scoring of these responses, namely assigning to them the lowest values of originality and feasibility among the responses given to the same target object in the group to which the participants belonged (always patients). Substantive differences observed in our results when using these two different scoring criteria are reported in footnotes.

3. Given the a-priori specification of directional hypotheses (schizophrenic outpatients were expected to perform worse than controls on all the creativity, cognitive, and affective measures) these tests were not adjusted for multiple comparisons (e.g., Pagano, 2013; Tucker, 1991).

4. Conclusions did not change when conducting Mann-Whitney tests, with the exceptions of the difference between groups in the RAT nonresponse (U = 81.5, Z = 2.15, p = .03) and in the AUT feasibility measure (U = 98.5, Z = −1.57, p = .12). When using the more lenient criterion for scoring originality and feasibility in the case of nonresponses, the difference between patients and healthy controls was still significant in peak originality, t(32) = −2.74, p < .01 (M_patients = 2.31, DS = .43; M_controls = 2.70, DS = .38), but not in average originality, t(32) = −0.83, p = .41 (M_patients = 2.09, DS = .16; M_controls = 2.12, DS = .07), and in feasibility, t(32) = −1.95, p = .06 (M_patients = 4.27, DS = .32; M_controls = 4.44, DS = .14).

5. With the more lenient scoring of patients’ nonresponses, the correlation coefficients were lower but still positive: AUT fluency with average originality (r = .29, p = .09), peak originality (r = .63, p < .01), and feasibility (r = .38, p < .05).

6. An additional set of analyses carried out without controlling for fluency indicated that the affective status did not account for the difference between patients and healthy controls in AUT average originality, peak originality, and feasibility. However, while controlling for the cognitive status did not account for the difference in peak originality, the differences in average originality and feasibility were no more significant when controlling for this factor (cf. Supplementary Materials). This seems to suggest that average originality and feasibility are more related to basic cognitive aspects than peak originality, but this conclusion has to be taken with caution considering that none of the covariates reached significance and that the differences between patients and controls were no more significant after controlling for AUT fluency.